Is Velocity Really the Golden Measurement?

Posted by Mike Bria on Jan 21, 2008 05:28 PM

Community

Agile

Topics

Delivering Value

Tags

XP ,

Retrospectives ,

Continuous Improvement ,

Value & Metrics ,

Planning

Recognized XP thought-leader, consultant, and author J.B. Rainsberger asks teams to consider if they may be giving agile development’s golden metric – velocity – a bit more credit than it deserves. In a recent article Rainsberger presents his case on velocity and effective agile team measurement, illustrating it in terms of home budgeting.

Think about budgeting your expenses at home: you know how much you actually pay for some expenses like car payments, mortgage or rent, insurance … there are other expenses that vary from month to month: food, clothing, entertainment, communications … [and] there are still other expenses that you just can’t foresee. You are injured in an accident and rack up $20k in hospital bills; this is the year out of the last ten that you finally need new shingles on the roof; you find out your son needs braces.

He observes this home budgeting truth -- that “there is no typical month” -- to be a lot like agile development and our use of velocity:

This is also why you cannot rely on velocity to plan far ahead: there is no such thing as a typical project or even a typical iteration.

This being the case, might it not be important for teams to spend less energy on scrutinizing velocity, and more time thoughtfully identifying areas of waste the team can benefit most from by eliminating? J.B. describes an experience with this, again in 'personal finance' terms, as follows:

We measured our actual spending, we looked at those numbers to identify waste, then we eliminated it. We did not waste effort staring at our budget [or, in agile terms, velocity], wondering how accurate it was, wondering how realistic it was, desperately trying to spread not enough money over too many expenses.

Basically, J.B. encourages teams to trade off some of the energy they spend contemplating velocity to focus more on asking questions that enable them to measure, pick, and eliminate the things they spend time on that no-one actually values. Said another way, the message here can be interpreted as a reminder for teams to make sure they’re being diligent about including good retrospectives as part of their normal iteration cycle.

Click the green tags below for more on the closely related InfoQ topics of retrospectives, continuous improvement, and kaizen.

1. Back to top

   Velocity measurement is not that useless either

   Jan 23, 2008 12:09 PM by Vikas Hazrati

   I would agree with the author that Velocity might not be the best measurement however on a tangent I believe that the velocity figures over a few iterations let you focus your energies in the right direction.
   For example consider this
   
   There is a team of 6 members which has velocity figures of
   It 1 - 12
   It 2 - 10
   It 3 - 9
   It 4 -7
   
   then looking at the figures iteration on iteration, you would try to do a deeper analysis to understand what is wrong. This would be similar to the news story www.infoq.com/news/2007/12/static-code-analysis... wherein the velocity would give you just visible information whereas a deeper analysis would bring out the root cause.
   
   Vikas Hazrati
   vikashazrati.wordpress.com

   Reply

2. Back to top

   Re: Velocity measurement is not that useless either

   Jan 23, 2008 2:54 PM by Mike Bria

   Vikas --
   
   Thanks for the response, I fully agree!
   
   Cheers!
   --MB

   Reply

3. Back to top

   Re: Velocity measurement is not that useless either

   Jan 24, 2008 1:51 PM by J. B. Rainsberger

   I agree that velocity itself is not a problem; however, when we measure velocity without analyzing how to eliminate waste, the impact is largely negative: we reduce the discussion to "how to go faster", as if typing speed were the bottleneck. Unfortunately, not enough teams reach for the deeper analysis, and if a team chooses not to analyze their situation more deeply, then I'm not sure they should measure velocity at all.

   Reply

• Older >